The Use of Ecological Niche Modeling to Infer Envenomation Risk of Apistobuthus Susanae Lourenço, 1998 (Arachnida: Scorpiones) in Southern Iran
Total Page:16
File Type:pdf, Size:1020Kb
BIHAREAN BIOLOGIST 15 (1): 1-5 ©Biharean Biologist, Oradea, Romania, 2021 Article No.: e201206 http://biozoojournals.ro/bihbiol/index.html The use of ecological niche modeling to infer envenomation risk of Apistobuthus susanae Lourenço, 1998 (Arachnida: Scorpiones) in Southern Iran Seyed Mahdi KAZEMI1, Mahboubeh Sadat HOSSEINZADEH2 and Kerim ÇIÇEK3 1. Zagros Herpetological Institute, 37156-88415, P. O. No 12, Somayyeh 14 Avenue, Qom, Iran. 2. Department of Biology, Faculty of Science, University of Birjand, Birjand, Iran. 3. Zoology Section, Department of Biology, Faculty of Science, Ege University, 35100, Izmir, Turkey. * Corresponding author, S.M. Kazemi, E-mail: [email protected] Received: 30. May 2020 / Accepted: 23. July 2020 / Available online: 20. July 2020 / Printed: June 2021 Abstract. The ability to identify the spatial distribution of medically important scorpion species is crucial for public health and conservation management. Scorpion stings are one of the important public health problems in south and southwest Iran. A potentially valuable approach is maximum entropy (MaxEnt) spatial distribution modeling which was applied here to digitize a map of potential distribution of Apistobuthus susanae. To achieve modeling distribution, seven bioclimatic variables with 11 spatially well-dispersed species occurrence records were used in the analysis. The most important environmental variable is precipitation seasonality that affected 79% of the distribution of A. susanae. Overall, the model could provide valuable help in searching for suitable habitat where it is hitherto unknown and also to manage health care and conservation activities. Key words: ecological niche modeling, envenomation risk, potential distribution, scorpion, Middle East. Introduction Apistobuthus (one doubtful specimen in Kuwait). The habitat of Apistobuthus susanae encompasses desert sites having hot Scorpions are a distinctive group of arthropods which di- and sandy conditions in Khuzestan and Ilam at elevations verged from other arachnids about 440 million years ago less than 130 m (Navidpour et al. 2008a,b, Navidpour & (Dunlop & Selden 2013, Waddington et al. 2015). Totally, 78 Lowe 2009, Figure 1). species of scorpions belonging to 20 genera and four families In the old world, Iran is acknowledged as one of the have been identified in Iran. Fifty out of 78 species of the world’s hotspots for scorpions (Ward et al. 2018). Annually, Iranian scorpions are endemic to Iran (Cokendolpher et al. more than 42,500 scorpion stings from 2001 to 2009 have 2019, Kazemi & Sabatier 2019, Kovařík 2019, Kovařík & been reported with about a 19.5% fatality rate. Hemiscorpius Navidpour 2020, Kovařík et al. 2020). Hence, Iran is lepturus, Androctonus crassicauda, Mesobuthus eupeus, Odonto- acknowledged as one of the world's hotspots for scorpions buthus doriae, Hottentotta saulcyi, Hottentotta schach, Compsobu- (Ward et al. 2018, Kazemi & Sabatier 2019). The Iranian thus matthiesseni, Olivierus caucasicus, Orthochirus scrobicu- scorpion fauna, 10 species (Hemiscorpius lepturus, Androcto- losus, and Apistobuthus pterygocercus are significant species in nus crassicauda, Apistobuthus susanae, Hottentotta saulcyi, H. terms of medical and pharmacological relevance (Dehghani zagrosensis, Compsobuthus matthiesseni, Mesobuthus phillipsii, & Fathi 2012). Odontobuthus doriae, Olivierus caucasicus and Orthochirus scro- Scorpion envenomation is a major public health problem biculosus) are medically important and implicated in enven- in developing countries, especially in tropical and subtropi- oming humans (Dehghani & Fathi 2012, Jalali & Rahim 2014, cal regions (Mazzei de Davila et al. 1997, Radmanesh Kazemi & Sabatier 2019). Based on Hauke & Herzig (2017) 1990a,b, 1998, Ghalim et al. 2000, Mahadevan 2000, Osnaya- and Ward et al. (2018), the genera of Androctonus, Apistobu- Romero et al. 2001, Ozkan et al. 2006). The highest percent- thus, Buthacus, Compsobuthus, Hottentotta, Mesobuthus, Odon- ages of scorpion stings were reported in Khuzestan, Koh- tobuthus, Orthochirus and Hemiscorpius contain dangerous giluye and Boyer-Ahmad and Ilam (Azhang & Moghisi 2006, and medically important species in Iran. Kazemi & Sabatier Shahi et al. 2015). In southern regions of Iran, especially in (2019) mentioned that A. crassicauda and H. lepturus have the Khuzestan Province, as in numerous tropical countries, en- highest risk of envenoming humans. venomation by scorpion stings is a major public health prob- The genus Apistobuthus Finnegan, 1932 includes two spe- lem. Khuzestan Province, neighboring with Iraq in the west cies, A. pterygocercus Finnegan 1932 and A. susanae Lourenço, and Persian Gulf in the south has mild winters, with average 1998 (Farzanpay 1987, Navidpour & Lowe 2009, Mirshamsi January temperatures ranging from 7o C to 18 o C, very hu- et al. 2011) belonging to family Buthidae Koch 1837; this is a mid and hot in summer, with temperature exceeding 48 o C significant genus in terms of medical and pharmacological during July in the interior areas (Shahbazzadeh et al. 2009). relevance in the Middle East (Al-Sadoon & Jarrar 2003, In addition, a scorpion sting is a main public health issue Dehghani & Fathi 2012, Jalali & Rahim 2014, Hauke & Her- in the region, especially for children and young adolescents zig 2017, Ward et al. 2018, Kazemi & Sabatier 2019). (Mirdehghan & Motlagh 2001). Due to their geographical The genus occurs in arid (hot, lowland sandy areas) and position and climate, scorpions had been a common danger semiarid regions. The genus distribution in terrestrial ecore- as reflected through scientific literature and religious texts of gions: Tigris-Euphrates alluvial salt marsh, Arabian Desert ancient Iran. According to the published reports of the Dis- and East Sahero-Arabian xeric shrublands and southern Iran ease Management Center of Iran, the number of scorpion Nubo-Sindian desert and semi-desert. Lourenço (1998), cor- stings was between 37,535 to 42,085 in Iran from 2001 to 2005 rectly recognized the single type specimen from Ahvaz whereas 24, 14, 23, 29 and 14 cases of these resulted in death, (Khuzestan Province; Iran) as belonging to a new species of respectively (Azhang & Moghisi 2006). 2 Kazemi S.M. et al. However, ecological niche modeling estimates the rela- given cell on the basis of environmental features in the same cell. tionship between environmental variables at species occur- MaxEnt is an ecological niche modeling (ENM) technique for making rences and environmental characteristics within the species’ predictions or inferences from incomplete information, as instead of real absence data it uses background data points (Phillips et al. 2006). general area of occurrence (Franklin 2009). Hence, predictive We implemented the randomly selected background approach (Phil- models of potential geographic distributions are widely used lips et al. 2006) and k-1 jackknife method recommended for working for a variety of applications in ecology, conservation and bi- with relatively small data sets (Pearson et al. 2007, Shcheglovitova & ogeography (Graham et al. 2004, Guisan & Thuiller, 2005). Anderson 2013). We built models with regularization multiplier val- MaxEnt is one of the algorithms that can be used for the ues ranging from 0.5 to 10 (increments of 0.5) and with six different prediction of a species’ potential distribution. It is a ma- feature class combinations (L, LQ, H, LQH, and LQHP, where chine-learning approach evaluating the likelihood of pres- L=linear, Q=quadratic, H=hinge, P=product and T=threshold), re- ence in a given cell on the basis of environmental features in sulting in 100 individual model runs. Model accuracy was evaluated from four evaluation metrics by the same cell; it is based on presence-only data (Franklin ENMs (Muscarella et al. 2014) in R vers. 3.5.2: the area under the 1995, Guisan & Thuiller 2005, Elith et al. 2006, Wisz et al. curve of the receiver operating characteristic plot for test localities 2008, Elith et al. 2010, Elith et al. 2011). (AUCTEST) (Hanley & Mcneil 1982, Peterson et al. 2011), the differ- Apistobuthus susanae is more dangerous than other scor- ence between training and testing AUC (AUCDIFF) (Warren & Seifert pions in Iran and information on its distribution and habitat 2011), OR10 (10% training omission rate) for test localities (Fielding preferences is also more limited than for other scorpion spe- & Bell 1997, Peterson et al. 2011) and the Akaike information criteri- cies. In this study, we predicted the potential distribution of on corrected for small sample sizes (AICc) (Burnham & Anderson 2004, Warren & Seifert 2011). A. susanae based on its realized niche. Modelling the poten- We applied the maximum test sensitivity plus specificity ap- tial distribution of A. susanae may shed more light on habitat proach as recommended by Liu et al. (2005) to transform the cloglog preferences and may suggest further potentially suitable output into a continuous map of the presence-absence distribution. habitats harboring so far undetected populations. The cloglog outputs represent habitat suitability from 0 (unsuitable) to 1 (suitable). Results were imported and visualized with ArcGIS v10.7. Material and Methods A total of 11 occurrence records of A. susanae were compiled from Results our own field work and literature sources (Navidpour et al. 2008a,b). The present records belonging to southwestern Iran include Model performances was high with average AUC = Khuzestan, Ilam Kohgiluyeh and Boyer-Ahmad Provinces.